Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add ai4s-research/ai4s-skills --skill paper-writergit clone --depth 1 https://github.com/ai4s-research/ai4s-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/ai4s-research/ai4s-skills/paper-writer)<a href="https://agentmods.dev/skills/ai4s-research/ai4s-skills/paper-writer"><img src="https://agentmods.dev/badge/skills/ai4s-research/ai4s-skills/paper-writer/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/ai4s-research/ai4s-skills/paper-writer"><img src="https://agentmods.dev/badge/skills/ai4s-research/ai4s-skills/paper-writer.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00054 | $0.02305 |
| Opus 5.5 | $0.00022 | $0.00922 |
| Sonnet 5.5 | $0.00011 | $0.00461 |
| Haiku 4.5 | $0.00005 | $0.00231 |
Grade A, and why
paper-writer scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 168 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Paper Writer
Overview
End-to-end research paper builder. Single stage, full quality from the start — there is no skeleton phase to enrich later. The agent (Claude Code / Cursor / Aider / Codex / …) does the writing using its own tools (WebFetch, WebSearch, Write, Bash). This skill has no Python runtime; it is purely a procedure + reference playbooks + a LaTeX template.
The substantive work is decomposed into reference playbooks under references/:
| Reference | Topic |
|---|---|
references/00-incremental-execution.md |
how to actually do this without losing work: batch sizes, persistence, resume — read first |
references/01-bibliography-expansion.md |
grow bibliography.bib to 200+ real entries via WebFetch/WebSearch |
references/02-figures-publication-grade.md |
TikZ / matplotlib / seaborn / multi-panel figure recipes |
references/03-section-playbook.md |
per-section structure, length, citation density |
references/04-layout-discipline.md |
tables, figures, floats, cross-refs, author + disclosure footnote |
references/05-quality-gate.md |
self-check before delivery (G1–G8 hard, S1–S4 soft) |
references/06-experiment-provenance.md |
honest provenance for every number (measured / simulated / illustrative) |
Read the relevant reference before writing, not after.
The full pass does not fit in a single turn. The bibliography is built across ~20+ small WebFetch/WebSearch batches; sections are drafted one per turn; figures are generated one at a time. Read references/00-incremental-execution.md before starting — it is the only execution mode that actually completes without losing work.
When to Use
- User asks to "write a paper" on a specific topic.
- User wants Abstract + Introduction + Related Work + Method + Experiment + Results + Conclusion.
- User has experiment results (a
results.json) and wants them formatted into a paper.
When NOT to Use
- User wants only a literature survey → the
literature-surveyskill. - User wants only the experiment package → the
experiment-suiteskill. - User wants only direction/topic exploration → the
research-explorerskill. - User wants the full multi-skill pipeline → the
ai4s-agentskill (which invokes this skill as one stage).
What ships with it
12 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- references/00-incremental-execution.md 11 KB
- references/01-bibliography-expansion.md 10 KB
- references/02-figures-publication-grade.md 20 KB
- references/03-section-playbook.md 13 KB
- references/04-layout-discipline.md 14 KB
- references/05-quality-gate.md 15 KB
- references/06-experiment-provenance.md 4.6 KB
- templates/paper/check_bibliography_freshness.py 3.8 KB runs code
- templates/paper/compile.sh 613 B runs code
- templates/paper/figures/.gitkeep 0 B
- templates/paper/main.tex 2.1 KB
- templates/paper/sections/.gitkeep 0 B
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 168 lines · 54 tokens per session scan A fa9ba55e64e5
paper-writer is a skill published in the GitHub repository ai4s-research/ai4s-skills (256 stars, last pushed 3mo ago), licensed MIT. It adds 54 tokens to every session and 2,305 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
remote-compute
Use when the user asks to run, submit, monitor, or cancel a job on a remote machine over SSH — their own GPU/CPU server, a workstation, or a Slurm cluster ("the cluster", a login node, "my 3090 box", "the compute server"). Picks a saved machine, runs the work directly over SSH (or via Slurm when present), tracks it…
domain-check
Use ONLY when the user asks to check, validate, review or audit analysis code for scientific correctness, or when you are the reviewer agent running its checks. Never run it unprompted as part of ordinary analysis work. Covers physics, earth/geo, biology, chemistry, social science and bioprocess/fermentation. Runs a…
computer-use
Use when a task needs an app on the user's own screen — a native desktop application (instrument control software, Origin, GraphPad, ImageJ, a reference manager, a spreadsheet), or a desktop browser window that needs window-level control. Read before the first computer call. Covers the observe-act-observe loop, why…
publication-figures
Use whenever you generate or review a chart, plot, table, or paper figure in this workspace, including work delegated by paper-writing, literature-survey, and experiment skills. Applies the Open Science publication style, enforces readable final-size layout for figures and tables, and rejects generic diagram-tool…
large-file
Use BEFORE reading any data file that could be large (CSV/TSV, Parquet, HDF5, FITS, NetCDF, NDJSON, genomics FASTQ/FASTA/VCF/BAM, GRIB, ROOT, or big text/simulation logs like VASP OUTCAR). Returns a compact memory pointer — header/schema/shape/sample/key numbers — by introspection and sampling in bounded memory, so…
modal-run
Use when the user asks to run heavy or GPU work on Modal (the cloud compute platform) — writing a Modal function in the workspace, running it with the user's own modal CLI + token, and bringing results back. Data-to-compute for jobs too big for the laptop, without a Slurm cluster.